CPD3 data structures
The CPD2-style variable names are generally usable with CPD3 data, but much more control over the information associated with the variable is available with an expanded syntax. Here is the help information returned by calling da.get –help=variables.
The "variables" bare word argument sets the list of variables to read from the archive. This can be either a direct list of variables (such as "T_S11" or a alias for multiple variables, such as "S11a". In the single variable form regular expressions may be used, as long as they match the entire variable name. This list is delimited by commas. If a specification includes a colon then the part before the colon is treated as an override to the default archive. If it includes two colons then the first field is treated as the station, the second as the archive, and the third as the actual variable specification. If it includes four or more colons then the specification is treated as with three except that the trailing components specify the flavor (e.x. PM1 or PM10) restriction. This restriction consists of an optional prefix and the flavor name. If there is no prefix then the flavor is required. If the prefix is "!" then the flavor is excluded, that is the results will never include it. If the prefix is"=" then the flavor is added to the list of flavors that any match must have exactly. For example:"T_S11" specifies the variable in the "default" station (set either explicitly on the command line or inferred from the current directory) and the "default" archive (either the "raw" archive or the one set on the command line. "raw:T_S1[12]" specifies the variables T_S11 and T_S12 from the raw archive on the "default" station. "brw:raw:S11a" specifies the "S11a" alias record for the station "brw" and the "raw" archive. ":avgh:T_S11:pm1:!stddev" specifies T_S11 from the "default" station hourly averages restricted to only PM1 data, but not any calculated standard deviations. The string "everything" can also be used to retrieve all available variables.